Meta Develops AI Version of Zuckerberg for Internal Staff Use

Meta Platforms is building a photorealistic AI character modeled on Chief Executive Mark Zuckerberg, training the system on his voice, tone, and company views so staff can interact with a digital version of their boss.

Meta Platforms is building a photorealistic, AI-powered version of Chief Executive Mark Zuckerberg, designed to interact with the company’s nearly 79,000 employees, according to a Financial Times report.

The AI character is being trained on Zuckerberg’s mannerisms, tone, voice, and public statements, as well as his current positions on company strategy, the Financial Times reported. The stated goal is to help workers feel more connected to one of technology’s most prominent executives.

Zuckerberg, 41, is personally involved in training the AI character, per the Financial Times. It is unclear when the project will be made available to staff or what access controls will govern its use. Meta did not respond to requests for comment.

ALSO READ: London Fintech Round Raises $6M Seed to Automate Finance Teams

The AI Zuckerberg initiative is separate from a distinct internal tool first reported by the Wall Street Journal: a so-called CEO agent designed to give Zuckerberg faster access to internal company data and streamline his own workflow. Both projects sit within the company’s broader push to apply AI across its operations.

Meta’s recently established Superintelligence Labs is leading a broader effort to develop characters at the company. That group released its first large language model, Muse Spark, earlier this month. Meta stock rose approximately 7% following that announcement, per financial disclosures tracked by Yahoo Finance.

The Zuckerberg clone is not the company’s first experiment with AI-driven personas. Meta debuted celebrity-based chatbots in September 2023, including one modeled on rapper Snoop Dogg, who licensed his voice and likeness for the effort. A subsequent platform, AI Studio, allowed creators and ordinary users to build AI versions of themselves. That program drew scrutiny from regulators and child safety advocates after users created sexually explicit personas. Meta moved in January to block teenagers from accessing its AI characters until the experience is updated 

Separately, Zuckerberg has been spending five to 10 hours per week writing code and attending engineering review sessions, the Financial Times reported, signaling a more hands-on role in the company’s AI development.

The project arrives as Meta accelerates spending on AI infrastructure. The company has said it plans to invest up to $135 billion in AI development this year as it competes with OpenAI and Google. Reality Labs, Meta’s virtual reality division, has posted cumulative losses of nearly $80 billion since 2020, according to company filings on the Securities and Exchange Commission’s EDGAR system.

Meta also faces unresolved legal exposure. A jury in New Mexico ordered the company last month to pay $375 million in civil penalties related to consumer safety failures on its platforms. A California court separately found that Meta had deliberately engineered Instagram to be addictive, resulting in harm to a young user, per court records.

It is unclear whether the AI Zuckerberg system will eventually be extended beyond internal use to creators or the public, though the Financial Times reported Meta believes the model could be replicated by influencers operating on its platforms.

Disclaimer: This report is based on publicly available information. Nervnow has not independently verified any claims.

MORE ON META AI 
CoreWeave and Meta Sign $21B AI Cloud Deal Through 2032
Meta Launches First AI Model From Restructured Superintelligence Unit
Nebius-Meta $27B Deal Highlights Big Push in AI Infrastructure Expansion